Artificial intelligence system for automatic tooth detection and numbering in the mixed dentition in CBCT

Area of Science:

• Dental imaging and diagnostics
• Artificial intelligence in healthcare
• Paediatric dentistry

Background:

• Efficient tooth identification and numbering are crucial for dental treatment planning.
• Traditional methods can be time-consuming and labor-intensive.
• Mixed dentition presents unique challenges in dental analysis.

Purpose of the Study:

• To assess the accuracy and effectiveness of an AI algorithm for automated tooth segmentation.
• To evaluate the application of AI in paediatric cone-beam computed tomography (CBCT) volumes.
• To determine the utility of AI in managing mixed dentition cases.

Main Methods:

• Utilized the nnU-Netv2 algorithm for automated tooth segmentation.
• Applied the AI model to CBCT scans of paediatric patients with mixed dentition.
• Evaluated segmentation performance based on accuracy metrics.

Main Results:

• The AI model demonstrated high effectiveness and accuracy in segmenting teeth.
• Automated segmentation significantly reduced the time and effort required for analysis.
• nnU-Netv2 proved capable of handling the complexities of mixed dentition.

Conclusions:

• AI, specifically nnU-Netv2, offers a powerful tool for automating tooth segmentation in paediatric dentistry.
• AI significantly enhances efficiency in dental treatment planning for mixed dentition.
• AI models are valuable assets for dentists, saving time and improving planning accuracy.